Connecting science, modelling capability and management of harmful cyanobacteria blooms

Transcription

1 5 th National Cyanobacteria Workshop Brisbane, September 2016 Connecting science, modelling capability and management of harmful cyanobacteria blooms David Hamilton

2 Acknowledgements Ministry of Business and Innovation: Lake Resilience Marsden Fund: Toxic in Crowds Hugo Borges Susie Wood Ian Hawes Dan Dietrich Jonathan Puddick

3 Complexity of models Van Nes, E. and M. Scheffer, A strategy to improve the contribution of complex simulation models to ecological theory. Ecological Modelling 185:

4 Factors affecting development of CyanoHABs Meteorology Catchment management & BMPs Sample capture Cell physiology Water chemistry Human Health Risk assessment Water policy Hydrodynamics

5 Models to test theories We have not yet come face-to-face with our failure to do effective ecological science. If we had effective, interesting ecological theories we could be testing them, and they would dictate the accuracy of our measurements. In the absence of theories, we can be as careless as we like Rigler, F.H. and R.H. Peters Science and Limnology. Excellence in Ecology Series. Olendorf, Germany

6 Outline Hydrodynamics and modelling Are we making the connections (to specialists in this area)? Emerging technology Remote sensing (in situ, aerial) Fine-scale sampling What are we missing Prioritising key processes

7 Interdisciplinary science to improve prediction of CyanoHABs: are we there yet? Without doubt, it would help if sessions at conferences were mixed rather than putting the modellers into their own room theliteracymuse.files.wordpress.com/2015/07/alone-in-classroomblog1.jpg Flynn KJ (2005) Castles built on sand: disfunctionality in plankton models and the inadequacy of dialogue between biologists and modellers. J Plankton Res. 27(12):

8 A common feature of blooms

9 Lake Rotoehu phycocyanin monitoring Shallow (max. depth 14 m) Polymictic Eutrophic Cyanobacterial blooms

10 Water temperature[ ( o C) o C] Water temperature [ degc ] Fluorescence [ RFUB ] Lake Rotoehu: temperature, chlorophyll and phycocyanin fluorescence Temperature 0.5 m, 12 m 0.5 m 3 m 7 m m Chlorophyll Phycocyanin May 20-May 27-May 03-Jun 10-Jun 17-Jun 24-Jun 01-Jul 08-Jul 15-Jul 13-May 20-May 27-May 03-Jun 10-Jun 17-Jun 24-Jun 01-Jul 08-Jul 15-Jul 14-May 18-May 22-May 26-May 30-May 03-Jun 07-Jun 11-Jun 15-Jun 19-Jun 2

11 Water temperature ( o C) Water temperature ( o C) Water temperature [ o C] Water temperature[ o C] Lake Rotoehu: High-frequency monitoring and modelling results - temperature Buoy; 0.5 m0.5 m Model; 0.5 m0.5 m May 18-May 22-May 26-May 30-May 03-Jun 07-Jun 11-Jun 15-Jun 19-Jun 23-Jun 27-Jun 01-Jul 05-Jul 09-Jul 13-Jul 17-Jul Buoy 10.5 m Model 10.5 m Buoy; 10.5 m Model; 10.5 m 14-May 18-May 22-May 26-May 30-May 03-Jun 07-Jun 11-Jun 15-Jun 19-Jun 23-Jun 27-Jun 01-Jul 05-Jul 09-Jul 13-Jul 17-Jul Time (14 May to 19 July)

12 Phycocyanin (rfu) Phycocyanin (rfu) Lake Rotoehu cyanobacteria: High-frequency surface measurements and modelling Buoy diatoms chlorophyll 25 Model chlorophyll chlorophyll Buoy phycocyanin Model phycocyanin Buoy; phycocyanin Model; cyanobacteria May 18-May 22-May 26-May 30-May 03-Jun 07-Jun 11-Jun 15-Jun 19-Jun 23-Jun 27-Jun 01-Jul 05-Jul 09-Jul 13-Jul 17-Jul Time (14 May to 19 July) Hamilton et al A Global Lake Ecological Observatory Network (GLEON) for synthesising high frequency. Inland Waters

13 Depth (m) Depth (m) Cyanobacteria chl (mg m -3 ) Temperature ( o C) Simulation of temperature and cyanobacteria Buoy diatoms Model chlorophyll May 18-May 22-May 26-May 30-May 03-Jun 07-Jun 11-Jun 15-Jun 19-Jun 23-Jun 27-Jun 01-Jul 05-Jul 09-Jul 13-Jul 17-Jul

14 Density change (kg m -3 h -1 ) Interactions of buoyancy and turbulence AM PM Irradiance ( mol photons m -2 s -1 ) Evening Wallace and Hamilton (1999) L&O

15 Depth (m) Interactions of buoyancy and turbulence AM PM Depth of peak biomass Time (total = 24 h) Euphotic depth Evening Rabouille et al. (2003) C.R. Biologies

16 Temperature ( o C) Shallow lakes can stratify strongly (e.g. diurnally) Lake Rotorua (z ~ 3 m) 32 AM PM Jan 17 Jan 18 Jan 19 Jan Evening

17 3-D modelling of cyanobacteria in a large, shallow Lake Rotorua (North Island)

18 Simple models or complex ones? in defence of big ugly models, Logan (1994) pointed out that complex systems need complex solutions Van Nes, E. and M. Scheffer, A strategy to improve the contribution of complex simulation models to ecological theory. Ecological Modelling 185:

19 Cells will aggregate at a critical ratio of diffusion to floating Time scale of mixing in surface layer (mixed layer depth 2 /eddy diffusivity) Time scale of cells floating (mixed layer depth/floating velocity Péclet number (mixed time scale/floating time scale)

20 Colony diameter Interactions of turbulence, colony size and buoyancy O Brien et al. (2003) Hydrobiologia Turbulent dissipation

21 Cryo-sampling of cyanobacteria Puddick et al. (2016) L&O Methods

22 On-Land Mesocosms

23 Use of remote sensing for bloom monitoring: a surface-temporal component Allan, M.G., Hamilton, D.P., Hicks, B.J. and Brabyn, L. 2011: Landsat remote sensing of chlorophyll a concentrations in central North Island lakes of New Zealand. International Journal of 32(7):

24 High spatial variability in a Microcystis bloom Robson, B. J. and Hamilton, D.P. 2003: Summer flow event induces a cyanobacterial bloom in a seasonal Western Australian estuary. Journal of Marine and Freshwater Research 54:

25 Depth (m) Distributions of fluorescence: cyanobacterial bloom vs DCM 0 Fluorescence [rfu] Fluorescence (rfu) :00:00 18:30:00 23:30: /03 21:00 25/03 01:00 25/03 05:00 25/03 08:00 25/03 12:00 25/03 16:00 Simmonds B, Wood SA, Özkundakci D, Hamilton DP Phytoplankton succession and the formation of a deep chlorophyll maximum in a hypertrophic volcanic lake. Hydrobiologia 745 (1):

26 Linking surface and depth observations across Rotorua lakes Fluorescence (-)

27 Total chlorophyll a (mg m -3 ) How good are our measurements for validating our model? Day of year Elliott, J.A. and S.J. Thackeray (2004). The simulation of phytoplankton in shallow and deep lakes using PROTECH. Ecological Modelling 178 (2004)

28 Spatial scale in in modelling cyanobacteria Highly heterogeneous horizontally and vertically ( 3D model application) Heterogeneous vertically, homogeneous horizontally ( 1D model application) Oliver, R., D. Hamilton, J. Brookes & R. Oliver i(n press).. Physiology, blooms and prediction of planktonic cyanobacteria. In: Whitton, B. The Ecology of Cyanobacteria (2 nd Edition).

29 Where do we need more work to enhance models? Spatial variability and toxicity 28 surface samples 3 h routine Cell enumeration Nutrients Total microcystin Extracellular microcystin DNA

30

31 Where do we need more work to enhance models? Benthic Microcystis Large quantities of Microcystis on sediment surface

32 Where do we need more work to enhance models? Competition competitive exclusion, N-fixation? Borges et al. (2016)

33 Where do we need more work to enhance models? Toxicity variability in time, species, strain

34 Where do we need more work to enhance models? Lake Rotorua modelling as a decision support tool ROTAN catchment model Lake model Climate model High frequency monitoring Inform

35 Major cyanobacteria management decisions are based around models Lake Rotoiti diversion wall Smith VH, Wood SA, McBride CG, Atalah J, Hamilton DP, Abell J Phosphorus and nitrogen loading restraints are essential for successful eutrophication control of Lake Rotorua, New Zealand. Inland Waters 6(2):

36 Conclusions Hydrodynamics and modelling We require further progress in the application of coupled hydrodynamicbloom models. Many blooms are driven in the short-term by favourable physical environments that allow cyanobacteria to dis-entrain from bulk water movement. Emerging technology Many of the tools to support detailed understanding of in situ cyanobacteria distributions exist currently, at various scales (e.g., remote sensing (in situ, aerial) and cryo-sampling). Integration and crossreferencing of these tools is required. What are we missing Major gaps in cyanobacteria modelling include: life cycle models (particularly given the importance of benthic recruitment), colony formation and scale analysis to distinguish critical turbulence levels.